An anamoly in PI Speed Control

Are you sure with the cost "minimization " function implementation ...
Also what about current estimation/prediction ISP1 ....

Looks like something is missed in the application code...
 
Are you sure with the cost "minimization " function implementation ...
Also what about current estimation/prediction ISP1 ....

Looks like something is missed in the application code...
Yes, I'm confident about that. It's just difference between predicted and reference current. I'll have a variable and store the least calculated error. If any further states give lesser error, then that will be the new optimum.

i_kp is used instead of Isp1. That's it. Check the latest images I had sent for full MPCC code.
 
increase the P term to make it unstable....maybe it is just a tunning issue...but I doubt it you seem to have the speed loop tuned. I...if you can get it oscillating above and below the set point then probably just a tunning issue.

are you using the cross product for the torque calculation?

what does Ia Ib Ic look like...I've got an algorithm that uses DTC ...everything looks wonderful but when you eyeball the current it looks awful, which coincidentally ties into the torque namely the two zero vectors distorting the motor model.

I'm not sure how you are loading the motor....but generated electrical torque should be more than your set point...there is no point in a high torque value if you don't have enough load...easily proved just lower your set point..it looks like you have some load and this should be observable by a change in current, assuming you have phase currents that look sinusoidal. if you don't have a sinusoidal-looking current signal on the motor side....well that's where the fun begins ...welcome to motor control.
 
Top